There are 52 human Adenoviruses which infect different human tissues and hundreds of adenoviruses that infect other species ranging from fish to primates. These viruses are highly efficient nanomachines that deliver their genomic payload to the nucleus within an hour of infection. As DNA viruses, they do not integrate into host DNA, they can be produced to high titers using established GMP protocols, and they have demonstrated safety in research and human gene therapy applications for the expression of ectopic genes. However, to date, their potential applications have been hindered by the almost exclusive use of one variety, Ad5 or an Ad2/5 chimera and the inability to engineer and combine multiple genetic modifications rapidly and systematically. Thus, there is a great need to extend the repertoire adenoviral vectors beyond that of Ad2/5 and to develop a technological platform that facilitates the rapid, de novo assembly of novel adenoviral genomes from component parts, allowing the systematic incorporation of multiple modifications and heterologous elements. Such a system would take advantage of the natural viral architecture, which is highly efficient in both delivering and expressing 36 genes (not including splice variants). The system could provide powerful diagnostic agents and therapeutic agents that incorporate multiplex and quantitative measurements of the pathway activities deregulated in different tumor samples.
The potential of adenoviral vectors in several applications is hindered by the ability to manipulate the 36 kb viral genome rapidly and systematically. Furthermore, the adenoviral vectors used in basic research, animal models, gene therapy and oncolytic therapy are limited to Adenovirus (Ad) serotypes 2 and 5. Ad2 and Ad5 were among the first to be discovered and, as such, there is a legacy of vectors/tools with which to manipulate their genomes, particularly in the E1 region. Ad2/5 Fiber proteins infect epithelial cells by binding to the receptor, CAR. Unfortunately, CAR is not expressed on all cell types and is downregulated on many metastases. Furthermore, approximately 80% of the human population has pre-existing neutralizing antibodies against Ad2/5, which together with off-target liver uptake and inflammation, limits systemic applications. Thus, the use of Ad2/5 vectors for gene delivery and cancer therapy is not necessarily an optimal choice, quite the contrary, but largely an accident of history.
Our ultimate goal is to engineer potent viral cancer therapies that not only undergo tumor selective lytic replication but which can be administered systemically in repeated rounds of treatment, avoid liver toxicity, efficiently target and cross the torturous tumor vasculature, infect cells via disparate receptors, generate a tumor bystander effect by localized expression of pro-drug activating enzymes/toxins within the tumor and which reawaken a beneficial host anti-tumor immune response. These are major challenges which are further compounded by the inability of human adenovirus to replicate in mice. This precludes the evaluation of human oncolytic viruses in immune competent genetically engineered mouse models of cancer (GEMMs) which have many advantages over xenograft models.
There are 52 human adenoviruses, indicating highly specialized adaptation for infecting and replicating in different host tissue environments. Many of these viruses infect different tissues and have Fiber proteins that bind cellular receptors other than CAR as well as a distinct cohort of ‘E3’ immune-modulation genes. Their unique properties have not been extensively studied or exploited due to the lack of tools necessary to modify their genomes. Similarly, there are also adenoviruses that infect other species, including mouse adenovirus (MAV-1).
Provided herein are solutions to these and other problems in the art.